OpenAI launches GPT-6 Sol and Luna models

Summary

OpenAI expanded its GPT-6 series with updated Sol (for complex and coding tasks) and Luna (for high-volume administrative work) models. The company cut API prices to half of the previous 5.6 series and claims a roughly 50% reduction in hallucinations compared to prior generations.

Why it matters

This price cut, attributed to caching and inference improvements, lowers the barrier for developers relying on mid-tier models for production workloads. Anthropic's release of Opus 5.5 just 90 minutes earlier highlights how narrow the competitive window has become, forcing companies to reassess vendor dependency on a weekly basis.

Highlights

• GPT-6 Sol and Luna are offered via API at half the cost of their 5.6 series counterparts.
• OpenAI claims Sol achieved 'Astra-level reliability' in internal truthfulness tests at a fraction of the cost.
• Anthropic launched Opus 5.5 just 90 minutes before OpenAI's announcement, underscoring the pace of competition.

Critical take

All performance claims rest solely on OpenAI's internal evaluations and one-on-one comparisons with Anthropic, without independent benchmarks. The 90-minute gap between the two launches also raises questions about whether these releases stem from product maturity or competitive reflexes.